{"id":"arxiv/2606.09420","url":"https://arxiv.org/abs/2606.09420","tags":["math.OC","q-fin.PM"],"score":0,"title":"Benchmarking Deep Time Series Models for Equity Portfolios","domain":"math.OC","source":"arxiv","authors":["Aoxin Zhang","Yuhan Cheng","Kwanting Leung"],"reviews":0,"subject":"math.OC","summary":"Benchmarking forecasting architectures for daily equity portfolios is not just a prediction exercise. It also asks which model remains usable after preferences, costs, and portfolio constraints are imposed. We build a CRSP daily-stock benchmark for 15 deep and statistical time-series architectures over 2018--2024. The protocol combines common-window decile portfolios, stochastic multi-criteria acceptability analysis, a deployment-adjusted acceptability index, and a constrained quadratic portfolio layer with capacity, beta, industry, risk, leverage, and turnover controls. The index starts…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-06-08","reactions":{}}